Synthesis and Characterization of Perovskite-Type [K1−xNax]MgF3 Mixed Phases via the Fluorolytic Sol-Gel Synthesis

(1)

Article

Synthesis and Characterization of Perovskite-Type [K ₁ − x Na _x ]MgF ₃ Mixed Phases via the Fluorolytic Sol-Gel Synthesis

Florian Schütz, Linda Lange, Kerstin Scheurell, Gudrun Scholz and Erhard Kemnitz *

Department of Chemistry, Humboldt-Universität zu Berlin, 12489 Berlin, Germany; schuetzf@hu-berlin.de (F.S.);

lindalange92@emailn.de (L.L.); scheurek@rz.hu-berlin.de (K.S.); gudrun.scholz@chemie.hu-berlin.de (G.S.)

* Correspondence: erhard.kemnitz@chemie.hu-berlin.de; Tel.: +49-30-2093-7555 Received: 18 January 2018; Accepted: 27 January 2018; Published: 30 January 2018

Abstract:The focus of this article is the synthesis of perovskite-type [K1−xNax]MgF₃mixed phases via the room-temperature fluorolytic sol-gel approach. Different molar ratios of K/Na were examined and analyzed by¹⁹F MAS NMR and X-ray powder diffraction. Starting from pure KMgF3, a systematic substitution of potassium by sodium was evidenced when replacing K by Na. As long as the amount of sodium is less than 80% as compared to potassium, spectra just show [K4−xNaxF] environments in a [K1−xNax]MgF3mixed phase but separate structures appear when the amount of sodium is further increased. Moreover, colloidal dispersions of nanoscaled KMgF3particles were obtained, which were used to fabricate coatings on glass slides. Thin films showed antireflective behavior and high transmittance.

Keywords:nano; fluorperowskite; fluorolytic sol-gel synthesis; K for Na replacement

1. Introduction

In 2003, our group explored the fluorolytic sol-gel synthesis [1]. In contrast to the common aqueous sol-gel synthesis which leads to the formation of nanosized metal oxides, the fluorolytic sol-gel synthesis yields metal fluoride nanoparticles [2,3]. Because of the low refractive indexes of metal fluorides, obtained metal fluoride sols can be used to fabricate antireflective thin films on different substrates, e.g., glasses or polymers,viadip coating. Additionally, these sols can be subsequently processed to xerogels as well. This is achieved by evaporating the solvent and yields solid metal fluorides with high surface areas. Diversification of the general fluorolytic synthesis approach resulted in the access of a wide variety of fluoride based materials [4,5]. In addition, in this case, the powerful room temperature synthesis method is used to access homodispersed perovskite-type mixed phases of [K1−xNax]MgF3 on a nano-scaled dimension under ambient conditions. For many decades, scientists are dealing with perovskite-type compounds like KMgF3

and NaMgF₃for different purposes, especially because of their interesting physical properties [6,7], synthetic accessibility [8,9], and application-focused demands [10,11]. Furthermore, the synthesis of mixed phases was investigated and will be used for direct comparison [12]. In general, compounds like KMgF3act as high-performance ceramics and luminescence host matrices [13–15]. Hence, the synthetic access of nanosized [K1−xNax]MgF₃mixed phases via the fluorolytic sol-gel synthesis under ambient conditions, their characterization, and probing their applicability for antireflective coatings is the main focus of this article.

Crystals 2018,8, 66; doi:10.3390/cryst8020066 www.mdpi.com/journal/crystals

(2)

Crystals 2018,8, 66 2 of 9

2. Methods and Experimental

2.1. Methods

19F MAS NMR: ¹⁹F MAS NMR data were measured with a Bruker Advance 400 MHz spectrometer (Billerica, MA, USA). All dried powder samples were filled into 2.5 mm ZrO₂rotors and the main rotation frequency was applied at 20 kHz. Referred toδ= 0 ppm of CFCl3(usingα-AlF3as the secondary standard for calibration) and with a D1time of 5 s, every spectrum consists of 32 scans.

XRD:Powder diffraction analysis was done by a Seifert XRD 3003 TT (Schnaittach-Hormersdorf, Germany) with rotating sample holder using Cu radiation (Cu K_α1,2;λ= 1.54 Å; U = 40 kV; I = 40 mA;

Ni filter). All samples were measured in Bragg-Bretano geometry and results compared to powder diffraction files of COD-inorganic.

Dip-Coating:To fabricate thin films from nanoscopic dispersions, float glasses (40 mm×20 mm

×3 mm) were coated by dip coating. To ensure a layer thickness of d > 100 nm, the dip-coating process was repeated three times. Obtained layers were heated up to 450^◦C with a heating rate (HR) of 10 K·min⁻¹, and kept at this temperature for 15 min.

Transmission-Spectroscopy: Layer thickness (d), reflectance (R_min), transmission (Tmax), and refractive index n, using a FILMetrics Thin Film Analyzer F10-RT (Unterhaching, Germany) were determined. Spectra were measured in a wavelength range ofλ= 400–1050 nm.

2.2. Experimental Materials:

Compound Formula Concentration/Purity Supplier

Magnesium Ethoxide Mg(OEt)2 99.8% EVONIK Industries

Magnesium Chloride MgCl₂ anhydrous,≥98% Aldrich

Potassium Methoxide KOMe ≥95% Aldrich

Hydrogen Fluoride HF anhydrous, pure Solvay

Methanol MeOH ≥99.6% Aldrich

HF solution in Methanol HF_MeOH 20.53 mol·L⁻¹ -

Synthesis of HFMEOH:The alcoholic HF solution was prepared according to our previous report by dissolving HF in methanol [4].

Synthesis of KMgF3(coating sol):100 mL MeOH, 2.99 mL HF_MeOH(60 mmol), and 0.1 mL nitric acid (2 mmol) were added in a 250 mL PP bottle. In a schlenk flask, 1.40 g KOMe (20 mmol) and 2.29 g Mg(OEt)2(20 mmol) were dissolved in 90 mL MeOH. The grey solution of the potassium and magnesium precursors were transferred dropwise into the HF/HNO3solution under heavy stirring.

After one day, the grey dispersion turned into a milky sol and 0.3 mL TMOS (2 mmol) were added.

After 7 days of stirring, a clear, colorless sol was obtained.

Synthesis of NaMgF3: 100 mL MeOH, 2.99 mL HFMeOH (60 mmol), and 0.1 mL nitric acid (2 mmol) were added in a 250 mL PP bottle. In a schlenk flask, 0.46 g Na (20 mmol) and 2.29 g Mg(OEt)₂(20 mmol) were dissolved in 90 mL MeOH. The grey solution of the sodium and magnesium precursors were transferred dropwise into the HF/HNO3solution under heavy stirring. After one day, an additive (Table1) was added. After 7 days of stirring, a grey opaque dispersion was obtained.

Xerogel: 40 mL of NaMgF3sol were put into a glass flask. The solvent was removed under vacuum and the obtained powder was divided into three parts, whereas two were thermally treated at different temperatures.

Synthesis of [K1−xNax]MgF3:100 mL MeOH, 2.99 mL HFMeOH(60 mmol), and 0.1 mL nitric acid (2 mmol) were added in a 250 mL PP bottle. In a schlenk flask, Na (Table2), KOMe (Table2) and 2.29 g Mg(OEt)₂(20 mmol) were dissolved in 90 mL MeOH. The grey solution of the sodium/potassium and magnesium precursors were transferred dropwise into the HF/HNO3solution under heavy stirring.

(3)

After one day, an additive (Table1) was added. After 7 days of stirring, a grey opaque dispersion was obtained.

Xerogel:40 mL of [K1−xNax]MgF3sol were put into a glass flask. The solvent was removed under vacuum and the obtained powder was divided into three parts, whereas two were thermally treated at different temperatures.

Table 1.Amounts of Additives in [K_(1−x)Nax]MgF₃syntheses.

Additive n_Additive[mmol] v_Additive[mL]

TFA 2 0.15

Al(O^sBu)₃ 2 0.50

Zr(OⁱPr)4 2 0.62

TMOS 2 0.30

TFA: Trifluoro acetic acid, TMOS: Tetramethoxysilane.s: sec-,i: iso-.

Table 2.Amounts of reactants in [K(1−x)Nax]MgF₃syntheses.

Compound nNa[mmol] mNa[g] n_KOMe[mmol] m_KOMe[g]

[K_0.2Na_0.8]F₃ 16 0.37 4 0.28

[K_0.4Na_0.6]F₃ 12 0.28 8 0.56

[K_0.6Na_0.4]F₃ 8 0.18 12 0.84

[K_0.8Na_0.2]F₃ 4 0.09 16 1.12

3. Results and Discussion

[K1−xNax]MgF3phases of varying K to Na ratios were synthesized according the following general reaction stoichiometry (Scheme1).

Crystals 2018, 8, x FOR PEER REVIEW 3 of 9

Xerogel: 40 mL of [K1−xNax]MgF3 sol were put into a glass flask. The solvent was removed under vacuum and the obtained powder was divided into three parts, whereas two were thermally treated at different temperatures.

Table 1. Amounts of Additives in [K^(1−x)Na^x]MgF³syntheses.

Additive nAdditive [mmol] vAdditive [mL]

TFA 2 0.15

Al(O^sBu)3 2 0.50

Zr(OⁱPr)4 2 0.62

TMOS 2 0.30

TFA: Trifluoro acetic acid, TMOS: Tetramethoxysilane.

Table 2. Amounts of reactants in [K(1−x)Nax]MgF3 syntheses.

Compound nNa [mmol] mNa [g] nKOMe [mmol] mKOMe [g]

[K0.2Na0.8]F3 16 0.37 4 0.28 [K0.4Na0.6]F3 12 0.28 8 0.56 [K0.6Na0.4]F3 8 0.18 12 0.84 [K0.8Na0.2]F3 4 0.09 16 1.12

3. Results and Discussion

[K1−xNax]MgF3 phases of varying K to Na ratios were synthesized according the following general reaction stoichiometry (Scheme 1).

Scheme 1. Formation of [K1−xNax]MgF3 by fluorolytic sol-gel synthesis.

In addition to the reaction presented in the scheme above, Mg(OEt)2 was partly replaced by MgCl2 (10–30%).In analogy to our previous report, the idea was to generate a catalytic cycle in which HCl, formed by the reaction of MgCl2 with HF, would increase the reaction rate significantly due to the higher acidity of HCl as comapred to HF [4]. Unfortunately, the usage of MgCl2 within this reaction leads to the formation of [K/Na]Cl, and hence, is not recommended for this reaction system.

Thus, we tried to create a new catalytic cycle based on nitric acid: comparing the acidity of nitric acid sand hydrogen fluoride, nitric acid is a better proton donator towards metal–oxygen bonds than HF.

Fortunately, metal nitrates react subsequently with HF forming [K1−xNax]MgF3 mixed phases without any side products.

With the focus on easily access clear sols, which indicates the formation of mainly homodispersed nanoparticles inside the sol, we investigated the influence of temperature on clearing up rate during KMgF3 syntheses (Table 3). Thus, it turns out that clear sols can only be obtained by performing the synthesis under cooling. To ensure a long-time stability of the obtained sols, different additives were tested in order to avoid particle agglomeration and water induced gelation. Moreover, the addition of zeta-potential affecting electrolytes allows a faster clearing up of the KMgF3-sols.

Especially tetramethoxysilane (TMOS) and trifluoroacetic acid (TFA) turned out to cause a fast clearing up of the reaction systems resulting in very fast formation of clear sols. Comparing the visual appearance of different samples, the influence of TMOS and TFA is conspicuous (Figure 1). Based on dynamic light scattering (DLS) investigations, a mean particle size diameter of d = 69 nm within sol 12 was determined (Figure 2). It can be seen that the majority of the particles exhibit diameters less than 50 nm. Moreover, the correlation curve appears to be sigmoidal, which basically indicates a monodispers contribution.

Scheme 1.Formation of [K1−xNax]MgF₃by fluorolytic sol-gel synthesis.

In addition to the reaction presented in the scheme above, Mg(OEt)2was partly replaced by MgCl2(10–30%). In analogy to our previous report, the idea was to generate a catalytic cycle in which HCl, formed by the reaction of MgCl₂with HF, would increase the reaction rate significantly due to the higher acidity of HCl as comapred to HF [4]. Unfortunately, the usage of MgCl2within this reaction leads to the formation of [K/Na]Cl, and hence, is not recommended for this reaction system.

Thus, we tried to create a new catalytic cycle based on nitric acid: comparing the acidity of nitric acid sand hydrogen fluoride, nitric acid is a better proton donator towards metal–oxygen bonds than HF.

Fortunately, metal nitrates react subsequently with HF forming [K1−xNax]MgF3mixed phases without any side products.

With the focus on easily access clear sols, which indicates the formation of mainly homodispersed nanoparticles inside the sol, we investigated the influence of temperature on clearing up rate during KMgF3syntheses (Table3). Thus, it turns out that clear sols can only be obtained by performing the synthesis under cooling. To ensure a long-time stability of the obtained sols, different additives were tested in order to avoid particle agglomeration and water induced gelation. Moreover, the addition of zeta-potential affecting electrolytes allows a faster clearing up of the KMgF₃-sols. Especially tetramethoxysilane (TMOS) and trifluoroacetic acid (TFA) turned out to cause a fast clearing up of the reaction systems resulting in very fast formation of clear sols. Comparing the visual appearance of different samples, the influence of TMOS and TFA is conspicuous (Figure1). Based on dynamic light scattering (DLS) investigations, a mean particle size diameter of d = 69 nm within sol12was determined

(4)

Crystals 2018,8, 66 4 of 9

(Figure2). It can be seen that the majority of the particles exhibit diameters less than 50 nm. Moreover, the correlation curve appears to be sigmoidal, which basically indicates a monodispers contribution.

Table 3.Parameters of KMgF3syntheses.

# Concentration [mol/L] Temperature [^◦C] Additive Appearance

1 0.2 23 TFA opaque

2 Al(O^sBu)₃ opaque

3 Zr(OⁱPr)₄ opaque

4 TMOS slightly opaque

5 40 TFA opaque

7 Zr(OⁱPr)4 opaque

8 TMOS opaque

9 0 TFA slightly milky

12 TMOS clear

TFA: Trifluoroacetic acid, TMOS: Tetramethoxysilane.s: sec-,i: iso-.

Table 3. Parameters of KMgF³syntheses.

# Concentration [mol/L] Temperature [°C] Additive Appearance

1 0.2 23 TFA opaque

2 Al(O^sBu)3 opaque

5 40 TFA opaque

6 Al(O^sBu)3 opaque

8 TMOS opaque

10 Al(O^sBu)3 opaque

12 TMOS clear

TFA: Trifluoroacetic acid, TMOS: Tetramethoxysilane. s: sec-, i: iso-.

Figure 1. Comparison of visual appearances. (a) ethanol, (b) sol 12, (c) sol 9, (d) sol 10.

Figure 2. Volume weighted particle size distribution (Left) and corresponding correlation curve (Right) of sol 12.

To understand the action of the sol-stabilizing additives TMOS and TFA more precisely, IR- measurements of xerogel 9 and 12 (obtained from sol 9 and 12) were performed (Figure 3).

Comparing the IR spectra of obtained xerogels to pure additives, it can be seen that xerogel 9 shows similar stretching bands to pure TFA. Especially the shifting of the C=O band from1760 cm⁻¹(purple) to 1685 cm⁻¹(orange) supports the supposition of a coordinative bonding between the Lewis acidic sites of the nanoparticle and the trifluoro acetic acid molecules. Thus, electrons of the C=O HOMO drain off, leading to a weaker bonding and a red-shifted band within the IR spectrum. In contrast to this, no traces of TMOS within xerogel 12 were found, which may be taken as an indication for the zeta-potential affecting properties of TMOS.

Figure 1.Comparison of visual appearances. (a) ethanol, (b) sol12, (c) sol9, (d) sol10.

Table 3. Parameters of KMgF³syntheses.

# Concentration [mol/L] Temperature [°C] Additive Appearance

1 0.2 23 TFA opaque

2 Al(O^sBu)3 opaque

5 40 TFA opaque

6 Al(O^sBu)3 opaque

8 TMOS opaque

10 Al(O^sBu)3 opaque

12 TMOS clear

TFA: Trifluoroacetic acid, TMOS: Tetramethoxysilane. s: sec-, i: iso-.

Figure 1. Comparison of visual appearances. (a) ethanol, (b) sol 12, (c) sol 9, (d) sol 10.

Figure 2. Volume weighted particle size distribution (Left) and corresponding correlation curve (Right) of sol 12.

To understand the action of the sol-stabilizing additives TMOS and TFA more precisely, IR- measurements of xerogel 9 and 12 (obtained from sol 9 and 12) were performed (Figure 3).

Comparing the IR spectra of obtained xerogels to pure additives, it can be seen that xerogel 9 shows similar stretching bands to pure TFA. Especially the shifting of the C=O band from1760 cm⁻¹(purple) to 1685 cm⁻¹(orange) supports the supposition of a coordinative bonding between the Lewis acidic sites of the nanoparticle and the trifluoro acetic acid molecules. Thus, electrons of the C=O HOMO drain off, leading to a weaker bonding and a red-shifted band within the IR spectrum. In contrast to this, no traces of TMOS within xerogel 12 were found, which may be taken as an indication for the zeta-potential affecting properties of TMOS.

Figure 2.Volume weighted particle size distribution (Left) and corresponding correlation curve (Right) of sol12.

To understand the action of the sol-stabilizing additives TMOS and TFA more precisely, IR-measurements of xerogel 9 and 12 (obtained from sol 9 and 12) were performed (Figure 3).

Comparing the IR spectra of obtained xerogels to pure additives, it can be seen that xerogel9shows similar stretching bands to pure TFA. Especially the shifting of the C=O band from1760 cm⁻¹(purple) to 1685 cm⁻¹(orange) supports the supposition of a coordinative bonding between the Lewis acidic sites of the nanoparticle and the trifluoro acetic acid molecules. Thus, electrons of the C=O HOMO drain off, leading to a weaker bonding and a red-shifted band within the IR spectrum. In contrast to this, no traces of TMOS within xerogel12were found, which may be taken as an indication for the zeta-potential affecting properties of TMOS.

(5)

Figure 3. IR spectra of xerogel 9 (orange) and TFA (purple) (Left), IR spectra of xerogel 12 and TMOS (black) (Right).

To receive structural information, obtained X-ray amorphous xerogel 12 was thermally treated at 450 °C to increase crystallinity and to allow analysis by XRD. The ¹⁹F MAS NMR spectra of non- annealed sample of xerogel 12 show the main signal of KMgF³ at −184.6 ppm (Figure 4) [12].

Moreover, a small signal for MgF2 can be found at −198.2 ppm [4]. Thermally induced changes in the structure were not confirmed by comparing ¹⁹F MAS NMR spectra, but by XRD. While the non- annealed perovskite-type mixed phases of [K1−xNax]MgF3. xerogel is totally amorphous, the annealed xerogel shows good agreement with the powder diffraction file of KMgF3 (Figure 5, pdf: 18-1033, star). Hence, just amorphous KMgF3 can evidently be obtained not only via hydrothermal synthesis as described in literature but also via the fluorolytic sol-gel synthesis very easily under ambient conditions [9].

Figure 4. ¹⁹F MAS NMR of non-annealed xerogel 12, star: rotational side band, υ^rot = 20 kHz.

Figure 5. XRD of annealed (red) and non-annealed (black) xerogel 12, dashed lines: KMgF3 (pdf: 18- 1033, star).

Figure 3.IR spectra of xerogel9(orange) and TFA (purple) (Left), IR spectra of xerogel 12 and TMOS (black) (Right).

To receive structural information, obtained X-ray amorphous xerogel12was thermally treated at 450 ^◦C to increase crystallinity and to allow analysis by XRD. The ¹⁹F MAS NMR spectra of non-annealed sample of xerogel12show the main signal of KMgF₃at−184.6 ppm (Figure4) [12].

Moreover, a small signal for MgF2can be found at−198.2 ppm [4]. Thermally induced changes in the structure were not confirmed by comparing ¹⁹F MAS NMR spectra, but by XRD. While the non-annealed perovskite-type mixed phases of [K1−xNax]MgF3. xerogel is totally amorphous, the annealed xerogel shows good agreement with the powder diffraction file of KMgF₃(Figure5, pdf: 18-1033, star). Hence, just amorphous KMgF3can evidently be obtained not only via hydrothermal synthesis as described in literature but also via the fluorolytic sol-gel synthesis very easily under ambient conditions [9].

To receive structural information, obtained X-ray amorphous xerogel 12 was thermally treated at 450 °C to increase crystallinity and to allow analysis by XRD. The ¹⁹F MAS NMR spectra of non- annealed sample of xerogel 12 show the main signal of KMgF3 at −184.6 ppm (Figure 4) [12].

Figure 4. ¹⁹F MAS NMR of non-annealed xerogel 12, star: rotational side band, υ^rot = 20 kHz.

Figure 5. XRD of annealed (red) and non-annealed (black) xerogel 12, dashed lines: KMgF3 (pdf: 18- 1033, star).

Figure 4.¹⁹F MAS NMR of non-annealed xerogel12, star: rotational side band,υrot= 20 kHz.

To receive structural information, obtained X-ray amorphous xerogel 12 was thermally treated at 450 °C to increase crystallinity and to allow analysis by XRD. The ¹⁹F MAS NMR spectra of non- annealed sample of xerogel 12 show the main signal of KMgF3 at −184.6 ppm (Figure 4) [12].

Figure 4. ¹⁹F MAS NMR of non-annealed xerogel 12, star: rotational side band, υrot = 20 kHz.

Figure 5. XRD of annealed (red) and non-annealed (black) xerogel 12, dashed lines: KMgF³(pdf: 18- 1033, star).

Figure 5. XRD of annealed (red) and non-annealed (black) xerogel 12, dashed lines: KMgF₃ (pdf: 18-1033, star).

(6)

Crystals 2018,8, 66 6 of 9

Based on the results described above, we tried to obtain pure NaMgF3 and [K1−xNax]MgF3

mixed phases the same way by cooling down the reaction mixtures and adding TMOS subsequently.

Unfortunately, all sols obtained this way were not fully clear but opaque (Table4).

Table 4.Parameters of NaMgF₃and [K(1−x)Nax]MgF₃syntheses.

# Compound Concentration [mol/L] Temperature [^◦C] Additive Appearance

13 NaMgF₃ 0.2 0 TMOS Opaque

14 [K0.2Na0.8]F3 Opaque

15 [K_0.4Na_0.6]F₃ Opaque

17 [K_0.8Na_0.2]F₃ Opaque

TMOS: Tetramethoxysilane.

To receive structural information of obtained [K(1−x)Nax]MgF3mixed solid phases,¹⁹F MAS NMR of annealed xerogels9, 13–17were compared (Figure6). In line with literature reports, the replacement of K-cations by Na-cations can be confirmed [12]. Up to an amount of 20% of Na-doping, the cation mixed perovskite phases remain in a KMgF3 structural environment; the formation of a separate NaMgF3phase can be excluded. As already mentioned before,¹⁹F MAS NMR of xerogel9shows the main signal of KMgF₃at−184.6 ppm. With increasing Na⁺doping concentration, the spectra show a high field shifted widening. At a Na⁺doping level of 40% (xerogel16) a signal at−202.9 ppm appears, pointing out the formation of [Na4F] moieties. While this signal increases with higher amounts of Na⁺within the samples (xerogel15and14), the corresponding KMgF3signal decreases until complete disappearance (xerogel13). With the help of the dmfit program, separate signals in the range of−180 ppm and−200 ppm can be determined and allow the assignment of different fluoride ion environments within a cubic KMgF3structure (Figure7) [16]. The dmfit simulation shows four almost equidistant (∆δ_ave= 4.3 ppm) signals with different intensities at invariable peak-widths (Table5). The equidistant shifts to a higher field can be explained by a step-by-step substitution of K⁺ by Na⁺. Although it can be assumed that a statistic substitution occurs, the experimentally determined proportion for end-member environments [K4F] and [Na4F] is favored, and therefore, results given in the literature can be confirmed.

Based on the results described above, we tried to obtain pure NaMgF3 and [K1−xNax]MgF3 mixed phases the same way by cooling down the reaction mixtures and adding TMOS subsequently.

Unfortunately, all sols obtained this way were not fully clear but opaque (Table 4).

Table 4. Parameters of NaMgF³and [K^(1−x)Na^x]MgF³ syntheses.

# Compound Concentration [mol/L] Temperature [°C] Additive Appearance

13 NaMgF3 0.2 0 TMOS Opaque

14 [K^0.2Na^0.8]F³ Opaque

TMOS: Tetramethoxysilane.

To receive structural information of obtained [K(1−x)Nax]MgF3 mixed solid phases, ¹⁹F MAS NMR of annealed xerogels 9, 13–17 were compared (Figure 6). In line with literature reports, the replacement of K-cations by Na-cations can be confirmed [12]. Up to an amount of 20% of Na-doping, the cation mixed perovskite phases remain in a KMgF3 structural environment; the formation of a separate NaMgF³ phase can be excluded. As already mentioned before, ¹⁹F MAS NMR of xerogel 9 shows the main signal of KMgF³ at −184.6 ppm. With increasing Na⁺ doping concentration, the spectra show a high field shifted widening. At a Na⁺ doping level of 40% (xerogel 16) a signal at −202.9 ppm appears, pointing out the formation of [Na4F] moieties. While this signal increases with higher amounts of Na⁺ within the samples (xerogel 15 and 14), the corresponding KMgF3 signal decreases until complete disappearance (xerogel 13). With the help of the dmfit program, separate signals in the range of −180 ppm and −200 ppm can be determined and allow the assignment of diﬀerent fluoride ion environments within a cubic KMgF³ structure (Figure 7) [16]. The dmfit simulation shows four almost equidistant (Δδ^ave = 4.3 ppm) signals with different intensities at invariable peak-widths (Table 5). The equidistant shifts to a higher field can be explained by a step-by-step substitution of K⁺ by Na⁺. Although it can be assumed that a statistic substitution occurs, the experimentally determined proportion for end-member environments [K4F] and [Na4F] is favored, and therefore, results given in the literature can be confirmed.

Figure 6. ¹⁹F MAS NMR of xerogels 9, 13–17, star: spinning side band, υrot = 20 kHz.

Figure 6.¹⁹F MAS NMR of xerogels9,13–17, star: spinning side band,υrot= 20 kHz.

(7)

Figure 7. ¹⁹F MAS NMR of xerogel 17, dmfit analysis and corresponding fluorine sites [16].

Table 5. ¹⁹F MAS NMR signals of xerogel 17 by dmfit simulation [16].

n(Na⁺) δ (¹⁹F)exp. [ppm] Δδexp. [ppm] Proportionexp. [%] Proportioncalc. [%] nexp(K⁺) ncalc. (K⁺)

0 −184.6 3.8 (0,1) 64.8 41.0 2.59 1.64

1 −188.4 4.6 (1,2) 24.4 41.0 0.73 1.23

2 −193.0 4.5 (2,3) 7.2 15.4 0.14 0.31

3 −197.5 - 3.6 2.6 0.04 0.03

4 - - 0 0.2 0 0

ntotal(K⁺) 3.50 3.21

ntotal(K⁺)/4 0.88 0.80

This cation-replacement can also be evidenced by XRD measurements complying with the law of Vegard (Figure 8) [17]. With increasing amounts of the smaller Na-cations (rNa+ = 153 pm, rK+ = 178 pm) [18], the cell volume decreases leading to right-side shifting of peaks in the diffractograms. In comparison to the ¹⁹F MAS NMR spectra, a pure NaMgF3 phase cannot be found at x = 0.4. This may confirm the formation of [Na4F] moieties within a [K(1−x)Nax]MgF3 solid solution, but does not indicate the formation of pure a NaMgF³ phase. The varying numbers for calculated and experimentally determined n^total(K⁺) are in line with this assumption (Table 5). Even at Na-doping level of 60%, reflections of a NaMgF3 phase are difficult to identify, whereas reflections of both compounds show up at Na-levels ≥80% [K0.2Na0.8]MgF3.

Figure 8. XRDs of annealed xerogels 9, 13–17, pdf^x=0: 18-1033, star, pdf^x=1: 81-952.

Figure 7.¹⁹F MAS NMR of xerogel17, dmfit analysis and corresponding fluorine sites [16].

Table 5.¹⁹F MAS NMR signals of xerogel17by dmfit simulation [16].

n(Na⁺) δ(¹⁹F)exp.[ppm] ∆δexp.[ppm] Proportionexp.[%] Proportion_calc.[%] nexp(K⁺) n_calc.(K⁺)

0 −184.6 3.8 (0,1) 64.8 41.0 2.59 1.64

1 −188.4 4.6 (1,2) 24.4 41.0 0.73 1.23

2 −193.0 4.5 (2,3) 7.2 15.4 0.14 0.31

3 −197.5 - 3.6 2.6 0.04 0.03

4 - - 0 0.2 0 0

n_total(K⁺) 3.50 3.21

n_total(K⁺)/4 0.88 0.80

This cation-replacement can also be evidenced by XRD measurements complying with the law of Vegard (Figure8) [17]. With increasing amounts of the smaller Na-cations (r_Na⁺= 153 pm, rK+= 178 pm) [18], the cell volume decreases leading to right-side shifting of peaks in the diffractograms. In comparison to the¹⁹F MAS NMR spectra, a pure NaMgF3phase cannot be found at x= 0.4. This may confirm the formation of [Na4F] moieties within a [K(1−x)Nax]MgF3solid solution, but does not indicate the formation of pure a NaMgF₃phase. The varying numbers for calculated and experimentally determined ntotal(K⁺) are in line with this assumption (Table5). Even at Na-doping level of 60%, reflections of a NaMgF3 phase are difficult to identify, whereas reflections of both compounds show up at Na-levels≥80% [K0.2Na0.8]MgF3.

Figure 7. ¹⁹F MAS NMR of xerogel 17, dmfit analysis and corresponding fluorine sites [16].

Table 5. ¹⁹F MAS NMR signals of xerogel 17 by dmfit simulation [16].

n(Na⁺) δ (¹⁹F)exp. [ppm] Δδexp. [ppm] Proportionexp. [%] Proportioncalc. [%] nexp(K⁺) ncalc. (K⁺)

0 −184.6 3.8 (0,1) 64.8 41.0 2.59 1.64

1 −188.4 4.6 (1,2) 24.4 41.0 0.73 1.23

2 −193.0 4.5 (2,3) 7.2 15.4 0.14 0.31

3 −197.5 - 3.6 2.6 0.04 0.03

4 - - 0 0.2 0 0

ntotal(K⁺) 3.50 3.21

ntotal(K⁺)/4 0.88 0.80

This cation-replacement can also be evidenced by XRD measurements complying with the law of Vegard (Figure 8) [17]. With increasing amounts of the smaller Na-cations (rNa+ = 153 pm, rK+ = 178 pm) [18], the cell volume decreases leading to right-side shifting of peaks in the diffractograms. In comparison to the ¹⁹F MAS NMR spectra, a pure NaMgF3 phase cannot be found at x = 0.4. This may confirm the formation of [Na4F] moieties within a [K(1−x)Nax]MgF3 solid solution, but does not indicate the formation of pure a NaMgF3 phase. The varying numbers for calculated and experimentally determined ntotal(K⁺) are in line with this assumption (Table 5). Even at Na-doping level of 60%, reflections of a NaMgF3 phase are difficult to identify, whereas reflections of both compounds show up at Na-levels ≥80% [K0.2Na0.8]MgF3.

Figure 8. XRDs of annealed xerogels 9, 13–17, pdf^x=0: 18-1033, star, pdf^x=1: 81-952.

Figure 8.XRDs of annealed xerogels9,13–17, pdf_x=0: 18-1033, star, pdf_x=1: 81-952.

(8)

Crystals 2018,8, 66 8 of 9

Concerning the fact that sol12represents a water clear colorless sol, it was used for dip coating on glass-slides. After dip coating, glass slides were thermally treated at 450^◦C for 15 min. The obtained thin films show nearly perfect antireflective behavior with a remaining reflectance of just R = 0.15%

and transmission as high as T = 96.1% (Figure9, Table6).

Concerning the fact that sol 12 represents a water clear colorless sol, it was used for dip coating on glass-slides. After dip coating, glass slides were thermally treated at 450 °C for 15 min. The obtained thin films show nearly perfect antireflective behavior with a remaining reflectance of just R

= 0.15% and transmission as high as T = 96.1% (Figure 9, Table 6).

Figure 9. Reflectance and transmission curves of KMgF³thin film.

Table 6. Overview of layer properties.

Layer T [°C] t [min] dT [°C/min] d [nm] R [%] T [%] n632,8nm

1 450 15 10 116.7 0.15 96.1 1.27

4. Summary

Employing the room-temperature fluorolytic sol-gel synthesis, the formation of KMgF3, NaMgF3, and [K1-xNax]MgF3 mixed phases was evidenced. However, only with pure KMgF3 colorless and water clear sols were they obtained when the synthesis was performed under ice bath cooling and with the addition of TMOS as a stabilizer. With the help of ¹⁹F MAS NMR and XRD, we obtained evidence for the consecutive occupancy of potassium by sodium sites within a cubic KMgF3 structure.

If the Na dopant concentration exceeds 20% in relation to K, the formation of [Na4F] environments takes place and can be observed in ¹⁹F MAS NMR. In addition, the substitution does not take place statistically, instead, end-membered environments are favored. The appearance of a pure NaMgF3

phase within samples with less than 80% Na⁺ can be denied as evidenced by XRD measurements.

Thin films, which were obtained by dip coating with clear KMgF3 sols showed very low remaining reflectance and high transmission.

Acknowledgments: The authors thank the research training network GRK 1582/2 “Fluorine as a Key Element”

of DFG (Deutsche Forschungsgemeinschaft) for funding. This project was partly also funded by the German Federal Ministry of Economics and Technology (Grant 03ET1235C).

Author Contributions: Florian Schütz wrote the paper and did the experiments with the help of Linda Lange.

Kerstin Scheurell and Gudrun Scholz analyzed, evaluated and discussed the NMR data. Erhard Kemnitz was responsible for coordinating the experiments and correcting the manuscript.

Conflicts of Interest: The authors declare no conflict of interest.

References

1. Kemnitz, E.; Groß, U.; Rüdiger, S.; Shekar, C.S. Amorphe metallfluoride mit außergewöhnlich großer spezifischer oberfläche. Angew. Chem. 2003, 115, 4383–4386.

2. Hench, L.L.; West, J.K. The sol-gel process. Chem. Rev. 1990, 90, 33–72.

3. Rudiger, S.; Kemnitz, E. The fluorolytic sol-gel route to metal fluorides-a versatile process opening a variety of application fields. Dalton Trans. 2008, 1117–1127. doi:10.1039/B716483A.

Figure 9.Reflectance and transmission curves of KMgF₃thin film.

Table 6.Overview of layer properties.

Layer T [^◦C] t [min] dT [^◦C/min] d [nm] R [%] T [%] n_632,8nm

1 450 15 10 116.7 0.15 96.1 1.27

4. Summary

Employing the room-temperature fluorolytic sol-gel synthesis, the formation of KMgF3, NaMgF3, and [K1-xNax]MgF3mixed phases was evidenced. However, only with pure KMgF3colorless and water clear sols were they obtained when the synthesis was performed under ice bath cooling and with the addition of TMOS as a stabilizer. With the help of¹⁹F MAS NMR and XRD, we obtained evidence for the consecutive occupancy of potassium by sodium sites within a cubic KMgF3structure.

If the Na dopant concentration exceeds 20% in relation to K, the formation of [Na4F] environments takes place and can be observed in¹⁹F MAS NMR. In addition, the substitution does not take place statistically, instead, end-membered environments are favored. The appearance of a pure NaMgF₃ phase within samples with less than 80% Na⁺can be denied as evidenced by XRD measurements.

Thin films, which were obtained by dip coating with clear KMgF3sols showed very low remaining reflectance and high transmission.

Acknowledgments:The authors thank the research training network GRK 1582/2 “Fluorine as a Key Element” of DFG (Deutsche Forschungsgemeinschaft) for funding. This project was partly also funded by the German Federal Ministry of Economics and Technology (Grant 03ET1235C).

Author Contributions:Florian Schütz wrote the paper and did the experiments with the help of Linda Lange.

Kerstin Scheurell and Gudrun Scholz analyzed, evaluated and discussed the NMR data. Erhard Kemnitz was responsible for coordinating the experiments and correcting the manuscript.

Conflicts of Interest:The authors declare no conflict of interest.

References

1. Kemnitz, E.; Groß, U.; Rüdiger, S.; Shekar, C.S. Amorphe metallfluoride mit außergewöhnlich großer spezifischer oberfläche.Angew. Chem.2003,115, 4383–4386. [CrossRef]

2. Hench, L.L.; West, J.K. The sol-gel process.Chem. Rev.1990,90, 33–72. [CrossRef]

3. Rudiger, S.; Kemnitz, E. The fluorolytic sol-gel route to metal fluorides-a versatile process opening a variety of application fields.Dalton Trans.2008, 1117–1127. [CrossRef] [PubMed]

(9)

4. Krahl, T.; Broßke, D.; Scheurell, K.; Lintner, B.; Kemnitz, E. Novel aspects in the chemistry of the non-aqueous fluorolytic sol–gel synthesis of nanoscaled homodisperse MgF2sols for antireflective coatings.J. Mater. Chem.

C Mater. Opt. Electron. Devices2016,4, 1454–1466. [CrossRef]

5. Rehmer, A.; Scheurell, K.; Kemnitz, E. Formation of nanoscopic CaF₂via a fluorolytic sol-gel process for antireflective coatings.J. Mater. Chem. C Mater. Opt. Electron. Devices2015,3, 1716–1723. [CrossRef]

6. Adamkoviˇcová, K.; Fellner, P.; Kosa, L.; Nerad, I.; Proks, I.; Streˇcko, J. Determination of the enthalpy of fusion of NaMgF₃and KMgF₃.Thermochim. Acta1994,242, 23–26. [CrossRef]

7. Sahnoun, M.; Zbiri, M.; Daul, C.; Khenata, R.; Baltache, H.; Driz, M. Full potential calculation of structural, electronic and optical properties of KMgF₃.Mater. Chem. Phys.2005,91, 185–191. [CrossRef]

8. Sevonkaev, I.; Goia, D.V.; Matijevi´c, E. Formation and structure of cubic particles of sodium magnesium fluoride (neighborite).J. Colloid Interface Sci.2008,317, 130–136. [CrossRef] [PubMed]

9. Zhao, C.; Feng, S.; Chao, Z.; Shi, C.; Xu, R.; Ni, J. Hydrothermal synthesis of the complex fluorides LiBaF₃ and KMgF3with perovskite structures under mild conditions.Chem. Commun.1996, 1641–1642. [CrossRef]

10. Dotzler, C.; Williams, G.V.M.; Rieser, U.; Edgar, A. Optically stimulated luminescence in NaMgF₃:Eu²⁺. Appl. Phys. Lett.2007,91, 121910. [CrossRef]

11. Wu, M.; Song, E.H.; Chen, Z.T.; Ding, S.; Ye, S.; Zhou, J.J.; Xu, S.Q.; Zhang, Q.Y. Single-band red upconversion luminescence of Yb³⁺–Er³⁺via nonequivalent substitution in perovskite KMgF₃nanocrystals.J. Mater. Chem.

C Mater. Opt. Electron. Devices2016,4, 1675–1684. [CrossRef]

12. Martin, C.D.; Chaudhur, S.; Grey, C.P.; Parise, J.B. Effect of a-site cation radius on ordering of bx6 octahedra in (K, Na) MgF₃perovskite.Am. Mineral.2005,90, 1522–1533. [CrossRef]

13. Bacci, C.; Fioravanti, S.; Furetta, C.; Missouri, M.; Ramogida, G.; Rossetti, R.; Sanipoli, C.; Scacco, A.

Photoluminescence and thermally stimulated luminescence in KMgF₃:Eu²⁺crystals.Radiat. Prot. Dosim.

1993,47, 277–280. [CrossRef]

14. Bhalla, A.S.; Guo, R.; Roy, R. The perovskite structure—A review of its role in ceramic science and technology.

Mater. Res. Innov.2000,4, 3–26. [CrossRef]

15. Furetta, C.; Bacci, C.; Rispoli, B.; Sanipoli, C.; Scacco, A. Luminescence and dosimetric performances of KMgF3crystals doped with metal impurity ions.Radiat. Prot. Dosim.1990,33, 107–110. [CrossRef]

16. Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calvé, S.; Alonso, B.; Durand, J.-O.; Bujoli, B.; Gan, Z.;

Hoatson, G. Modelling one- and two-dimensional solid-state NMR spectra.Magn. Reson. Chem.2002,40, 70–76. [CrossRef]

17. Vegard, L. Die konstitution der mischkristalle und die raumfüllung der atome. Z. Phys. 1921,5, 17–26.

[CrossRef]

18. Shannon, R.T.; Prewitt, C.T. Effective ionic radii in oxides and fluorides.Acta Crystallogr. B Struct. Crystallogr.

Crys. Chem.1969,25, 925–946. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Synthesis and Characterization of Perovskite-Type [K1−xNax]MgF3 Mixed Phases via the Fluorolytic Sol-Gel Synthesis

Synthesis and Characterization of Perovskite-Type [K 1 − x Na x ]MgF 3 Mixed Phases via the Fluorolytic Sol-Gel Synthesis

Synthesis and Characterization of Perovskite-Type [K ₁ − x Na _x ]MgF ₃ Mixed Phases via the Fluorolytic Sol-Gel Synthesis